890
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
4.2 Observations and results
In the 80-g tests the AC-14 anchor appeared to be a reasonable
stable anchor. This means that pulling the anchor with the
device described above, the anchor digs into the sand and does
not rotate or rotated partly (up to 90 degrees). This was different
for the 1-g test. In this test the anchor rotated 180 degrees
around its pulling axis in front of the berm. In the model anchor
the flukes were fixed (different from a real anchor). This means
that when the anchor rotates, the flukes are pointing upwards
and the anchor will not dig into the sand or the gravel berm. To
avoid that the rotation of the anchor dominates all results the
last test was performed with the anchor just in front of the berm
and it was pulled over a short distance only.
The measured force displacements of both the 80-g tests and
the 1-g tests are shown in Figure 11. The forces measured in the
80-g tests were divided by 80 to make them comparable with
the results of the 1-g tests. Perfect scaling would mean that the
80-g test is 80 times higher, see Table 1. Thus dividing this
force by 80 should result in the same value as the result of the 1-
g test; Figure 11 shows that this is not the case. The force
measured in the 1-g test is relatively higher.
0
1
2
3
4
5
6
7
8
Force (N) scaled to 1g
-800
-600
-400
-200
0
200
400
600
800
Displacement (mm)
Test A
Test B
Test C
Test 1 80g
Test 2 80 g
80-g tests
1-g tests
Figure 11. Force versus displacement for 1-g and 80-g (scaled, see text)
tests.
Due to the rotation of the anchor just in front of the berm in
2 of the 3 tests, there is only one measurement of the maximum
penetration of the anchor in the berm. This was on average 25.4
mm for the 80-g tests and 21.8 mm for the 1-g test. The
difference is visible on the pictures taken after the test. After a
80-g test, Figure 12 the anchor flukes are completely in the
berm (one fluke is visible in the picture but this is because the
gravel is taken away for the measurement of the position of the
fluke, the fluke in the upper part of the picture shows the
original situation). Figure 13 shows that after the last 1-g test
the flukes do not completely penetrate into the berm.
Figure 12. Position of anchor at the end of an 80-g test.
Figure 13. Position of anchor at the end of last 1-g test.
5 DISCUSSION
All results indicate that the soil and berm at the low stress levels
of a 1-g test behave relatively stronger and stiffer than at the
original stress level that is present during an 80-g test. If the
stresses are not properly scaled, but lower than in reality; the
soil behavior in a model test is stiffer and stronger than in the
prototype. This means that also for purely friction materials as
tested here, the proper representation of the stress-state is
important. To test the protection efficiency of a berm against
anchor dragging, 1:5 scale tests at 1-g are quite common.
Looking at the results of this research, it is very likely that the
results of these 1:5 scale model tests underestimate the
penetration depth of the anchor in prototype, which is the
primary objective of these tests, because that determines
whether or not a pipe line is sufficiently protected. At a scale
1:5 the error will be smaller than at the scale 1:80 tested here,
but can still be of importance.
6 CONCLUSIONS
Comparing the results of anchor tests at a scale 1:80 at the
original stress level in a centrifuge with the results of a further
identical 1:80 test at 1-g with thus a reduced stress level, led to
the following conclusions:
- The drag forces at 1-g are higher than 1/80 of the drag forces
at 80-g .
- The stability of the anchor is less during the 1-g tests.
The penetration depth is lower in a 1-g test (only one test result)
- Consequently the results indicate that in general a 1-g scale
model test underestimates the penetration depth of the anchor
and therefore overestimates the protection efficiency of the
berm.
7 ACKNOWLEDGEMENTS
The authors want to acknowledge Deltares for providing the
possibility to perform the 1-g tests and Thijs van Dijk, Frans
Kop, Jennifer Rietdijk and Ferry Schenkeveld for their
ontribution to these tests.
c
8 REFERENCES
Bezuijen A. (2011), Rigid body calculations (personal communication).
LeQin Wang, HongKiat Chia. (2010
) Optimization study of pipeline
rock armour protection design based on finit element analyses
.
Proceedings of the ASME 2010 29th International Conference on
Ocean, Offshore and Arctic Engineering OMAE2010
© 2010
ASME.
Van Lottum H., Luger H.J., Bezuijen A. (2010) Centrifuge anchor
dragging tests in sand and clay. Proc.
Physical Modelling in
Geotechnics – Springman, Laue & Seward (eds)© 2010 Taylor &
Francis Group, London, ISBN 978-0-415-59288-8
1063-1068.
